»Really crazy container troubleshooting stories«
2017-10-21, 07:45–08:25, Event Loft

In this talk, the presenter will share a few container troubleshooting stories that were encountered in the life of an infrastructure operator. The use cases are deliberately chosen to be a bit advanced and focused around exploring the inner workings of core libraries and kernel, to remind everyone that even the lowest level of modern systems need some love.

The talk will follow a hands-on agenda, interactively iterating over all the key points of the troubleshooting process, focusing on the different tools used and providing immediate value to the listener, who should be able to apply the various workflows to other scenarios.

Example use cases presented:

  • Troubleshooting resource isolation between containers
  • Tracing the root cause of a crashing containerized application
  • Monitoring memory and performance issues in containers